MIT SCALE RESEARCH REPORT

advertisement
MIT SCALE RESEARCH REPORT
The MIT Global Supply Chain and Logistics Excellence
(SCALE) Network is an international alliance of
leading-edge research and education centers, dedicated
to the development and dissemination of global
innovation in supply chain and logistics.
The Global SCALE Network allows faculty, researchers,
students, and affiliated companies from all six centers
around the world to pool their expertise and collaborate
on projects that will create supply chain and logistics
innovations with global applications.
This reprint is intended to communicate research results
of innovative supply chain research completed by
faculty, researchers, and students of the Global SCALE
Network, thereby contributing to the greater public
knowledge about supply chains.
For more information, contact
MIT Global SCALE Network
Postal Address:
Massachusetts Institute of Technology 77
Massachusetts Avenue, Cambridge, MA 02139 (USA)
Location:
Building E40, Room 267
1 Amherst St.
Access:
Telephone: +1 617-253-5320
Fax: +1 617-253-4560
Email: scale@mit.edu
Website: scale.mit.edu
Research Report: MISI-2014-1
Demand Forecasting for Perishable Commodities:
A Case Study of Inflight Food Demand for Low Cost Airline
Daniel. F. Liman
MITGlobalScaleNetwork
For full thesis version please contact:
Professor Shardul Phadnis
Director of Research
MISI
No. 2A, Persiaran Tebar Layar, Seksyen
U8, Bukit Jelutong, Shah Alam,
40150 Selangor, Malaysia.
Phone: +6 03 7841 4845
Email: sphadnis@misi.edu.my
MITGlobalScaleNetwork
Demand Forecasting for Perishable Commodities:
A Case Study of Inflight Food Demand for Low Cost Airline
By: Daniel. F. Liman
Thesis Advisor: Dr. Asad Ata
Summary: Inflight food sales have become an important revenue source for low cost
airlines where demand forecasting is play a critical role in profitability. By creating a new
and improved forecasting method, the research is aiming for reduced waste and lost sales,
thus will increase the business profitability.
Daniel holds a Bachelors in Ocean engineering from
Institut Teknologi Bandung and MSCM from Malaysia
Institute for Supply Chain Innovation, part of MIT global
Scale Network. Prior to joining MISI, Daniel work as an
field engineer for Hafar Daya Konstruksi, where he
handle engineering issue for pipeline construction and
rectification.
KEY INSIGHTS
1. The two channels of food service, Pre-Book Meal and Inflight sales, have
different demand pool. Thus capturing demand from both channels will
greatly increase the revenue.
2. The demand of inflight perishable goods is normally distributed and affected
by seasonality.
3. News vendor model provide a more robust solution with significantly less
waste.
Introduction & Background
“Low-cost Airlines” or “Budget Airlines”
is a term used to address Airlines that
generally have lower fares in exchange
with less comfort / services. This business
model start at 1971 when Southwest
airlines, an America based airlines,
introduce a concept that eliminate some of
the extra services in order to reduce cost
and then pass the saving to the customers.
A business model that targeting customers
from middle and lower class, boosting
their demand by reducing the overall price,
where they only need to pay for what they
need. This model is a perfect picture of
“lean” that only accommodate the need of
transportation and forsake the other
additional services. A business model that
has been proven throughout the globe,
Started by Ryanair in 1995 at Europe
region and finally move to Asia in 2001.
Fresh food is one of the comfort element
that get eliminated during the creation of
this “lean” business concept. It is a tricky
product with uncertain demand and zero
salvage value, making it hard for the
airlines to maintain stock. In the other hand
it will reduce the customer service level
when the product is out of stock when
customers expect to be able to buy it,
creating a fundamental supply chain
problem. Many airlines have stopped
selling this product to completely remove
the expectation or shift to snack product
which last longer and has higher salvage
value. However the high revenue and
profit margin of fresh food product keep
tempting some airlines to maintain the
services.
Customers, defined as passengers who
board the plane, is the major key element
that affect the demand. Wubben and
Wangenheim have tested the Preto/NBD
model, classification of customers based
on “hiatus heuristic”, to predict which
customer will be the best customers in the
future based on historical data, an “ideal”
way to predict the behavior of each
customer and the possible incoming
demand (Wübben & Wangenheim, 2008)
Sales for food service is done through two
different channel, known as Pre-Book
Meal (PBM) and inflight sales. Both
channels have different characteristic and
degree of profitability, breaking down the
characteristics, weaknesses and strengths
of each channel is important to understand
the research environment.
Research & Analysis
The inflight food industry market size is
estimated to be at $18 billion per annum
with predicted growth of 5% per year
(Jones, 2007), creating a huge profit
opportunity for low cost airlines. Recent
research of Novie Johan and Peter Jones
from University of Surrey has create a
“Meal Multiplier” value on 3 region, based
on Delphi method on major carrier
feedback in each region. However low cost
airlines is different from general airlines, it
separate the meal cost from the ticket,
creating more vulnerability to profit loss
from wastage.
Forecasting is a key factor to tackle this
problem, so far the partner company has
tried a wide spectrum of forecasting
method, moving average, practitioner view
and the current heuristic method. Goldstein
& Gigerenzer (2009), in their paper, “Fast
and frugal forecasting”, has done a
research about implication of stochastic
and heuristic forecast method in multiple
industry, with a positive result for heuristic
method (Goldsteina & Gigerenzer, 2009).
However even heuristic method can’t
produce a 100% accuracy forecast, it is
simply impossible for the current system to
get the perfect customer data for each
flight. After all the forecasting method
should be applicable instead of a theory
that exist and has a potential. This
condition raise the question of “What is the
best forecasting approach to help low cost
airlines to reduce waste without losing
sales?”
A probability density function (PDF) is
then built based on the ratio application on
ten months sales data on one sector, which
then become a platform for news-vendor
model. With a critical ratio of 0,6 that
come from the following equation.
The initial proposition is to bring back the
stochastic part to enhance the heuristic
method and fill the “gap” of information
that exist in the method. Theoretically it
will create a random model with less
variability by pulling out the certain factor
from the variability pool, overall the
research will be a deductive research that
will try to test and validate this theory
Where
The model
To comply with data availability, the
research will be done on specific sample
on one sector to control the flight duration,
passenger’s structure, and flight departure
time. A validity test using a different
sample with completely different
parameters will be done to justify the
model.
Figure 1: Seat to food ratio PDF – Based on 10
months sales data
Seat to food ratio is the bridge that connect
the heuristic and stochastic part. Which
defined as:
The model generates an equation that will
determine the take up amount (number of
products on aggregate level), based on the
real time data.
The ratio run under the assumption of
separated pool of demand between PBM
and inflight sales. In this assumption the
PBM has its exclusive demand pool that is
separated from inflight sales, which come
purely from the exposure of products.
Therefore the demand pool for inflight
sales can be defined as the total passenger
count – PBM.
A simulation on historical data is done to
test the equation. Using Microsoft Excel as
the platform, the take up amount of the
equation is compared with the actual sales,
with additional two sales for every real life
stock out event. The result shows a
significant improvement in waste area with
a little trade off from sales, the detail can
be observed at the table below.
Sales
Current
operation
New
model
Waste
Profit
10395
2668
5170
9819
1348
5352
Table 1: Simulation result
Seasonality effect
To tackle this parameter, 2 different
approaches will be used. The first
approach is by dividing the domain for the
news vendor model, each month of the
year have its own PDF which capture the
change of demand between months. The
second approach is by using moving
average method to normalize the forecast
result based on the seasonality.
Tested on the sample, the first approach
show a slight improvement from previous
news vendor model with high level of
applicability. The second approach in the
other hand show no improvement and can
be ruled off from consideration.
Test on different sector
The next question is “how consistent is the
model?” The answer lie on another test on
the model using a flight from another
sector with completely different parameter.
Parameters
CGK-KL
Flight duration
3 hours 50 1 hour
minutes
MorningEvening
noon
East
south
Flight
departure
Domain
direction
HKG- KL
created based on academic view which
means it might not align with the industry
goal. Although the optimum result will
yield more profit, industry tends to look at
other point in providing service such as
customer satisfaction or minimize waste.
The next figure show the tradeoff between
waste and sales, which will become
important when the company make a
policy to control the minimum sales or
maximum waste for the system. Sales and
waste has an exponential relation that
restraint the model to achieve higher
customer service level.
10500
10000
9500
9000
8500
0
1000
2000
3000
4000
Figure 2: Tradeoff between waste (x axis) and
profit (y axis)
The last part of sensitivity analysis shows
the correlation of change in mean of ratio
B, sales, waste, and expected profit. This
analysis becomes important to reveal the
real demand distribution of direct sales.
Since the operation run at 20% stock out
rate, the actual sales data can’t represent
the higher part of demand, a condition that
put the news vendor model into lower end
area.
Table 2: Flight parameters
Once again the model show a significant
improvement, with 37% less waste and
only 7,4% trade off from sales. Therefore
it can be concluded that the model is
consistent.
Sensitivity analysis
Sensitivity analysis is an important key in
the model application, the model was
Figure 3: Sensitivity analysis on ratio B mean
This sensitivity analysis will explore the
higher end of demand by increasing the
mean of the base PDF of the news vendor
model.
Conclusion
Based on the simulation of Seat to food
ratio, where the model show a better
performance under the assumption of
different demand pool for PBM and
inflight sales, the proposition is accepted
and therefore the demand pool of PBM and
inflight sales are redeemed different. This
concept is in alignment with current
forecasting method of the company, where
inflight sales is categorized into different
entity.
Furthermore, under the assumption of
proposition, the application of news vendor
model show an improvement in profit for
the company, even with data limitation. It
shows a 1,5 to 4% profit improvement in
historical data test. This reveals a potential
of better forecast application for the
operation with sufficient data support.
Application of the model, however, might
not align with the future policy of the
company. Although the model can offer a
significant improvement in term of
reducing waste within the current sales
level, it shows a weakness in achieving
higher service level. News-vendor model
generate a certain level of stock to
accommodate a certain level of probability,
it can’t specifically adapt for each flight
Download